Photosynthetic acclimation, leaf turnover and growth in tree seedlings suddenly exposed to gaps in Jamaican montane rainforest

Abstract Mature leaves of tree seedlings were exposed to high light in four experimental gaps in the Jamaican upper montane rainforest (UMRF). Two of the six species studied were light-demanders: Alchornea latifolia and Clethra occidentalis. Two were gap-favoured: Pittosporum undulatum (an invasive) and Palicourea alpina (a subcanopy shrub). One was intermediate: Hedyosmum arborescens, and one was shade-tolerant: Guarea glabra. After five months, the following significant changes occurred in shade leaves that were exposed to gaps (‘shade-to-gap’ leaves; values as % of those in the pre-gap shade): maximum rate of photosynthesis + 40% (Alchornea), +35% (Clethra), −34% (Pittosporum), +72% (Palicourea); dark respiration +120% (Alchornea), +140% (Clethra), +60% (Pittosporum), +233% (Palicourea), +175% (Hedyosmum), +100% (Guarea); leaf thickness +18% (Alchornea), +18% (Clethra), +14% (Palicourea); leaf mass per unit area +18% (Alchornea), +15% (Pittosporum). Leaves produced in the gaps were (as a percentage of total live leaf number) 74% (Alchornea), 71% (Clethra), 50% (Pittosporum), 71% (Palicourea), 62% (Hedyosmum) and 50% (Guarea). Photosynthetic rates of leaves produced in the gaps were 53–120% higher than ‘shade-to-gap’ leaves. Overall, shade leaves on the three native, more light-demanding species (Alchornea, Clethra and Palicourea) showed photosynthetic acclimation, while the more shade-tolerant species (Hedyosmum and Guarea and Pittosporum undulatum) showed little acclimation in shade-to-gap leaves.


Introduction
Seedlings in the understorey, when suddenly exposed to much higher light, will have 'shade leaves' that are not adapted to the new light climate.When shade-grown seedlings or saplings are suddenly exposed to such conditions, for example, in a treefall gap, the overall response will be to replace shade leaves with 'sun leaves', but for the first few months, pre-existing shade leaves may be the main site of photosynthesis (Naidu and DeLucia 1997).The differences between shade leaves and sun leaves are well known.They include higher maximum rates of photosynthesis (A max ), dark respiration (R d ), leaf thickness (LT), N content and leaf mass per unit area (LMA) (Sims and Pearcy 1991, Newell et al. 1993, García-Núñez et al. 1995, Naidu and DeLucia 1997, Wu et al. 2018).Acclimation of the plant to the changed environment leads eventually to lower leaf area ratios and associated changes in morphology (Popma andBongers 1988, 1991;Kitajima 1994;Claussen 1996; Bloor and Grubb 2004;Ishii and Ohsugi 2011).However, comparatively little is known about how the shade leaves adapt to sudden increases in light.
There have been many experiments where plants in pots grown in shade are suddenly exposed to gap light conditions, but very few where naturally occurring seedlings in forests are suddenly exposed.In some species, the shade leaves acclimate soon after the increased light is imposed (often after a short period of photoinhibition; Kamaluddin and Grace 1992, Lovelock et al. 1994, 1998, Huang et al. 2015).There are likely to be differences between natural seedlings in forests and seedlings grown in pots even if the same soil is used.For example, in tropical lowland rainforest in Australia, a comparison between pot-grown seedlings in 0.8% daylight with seedlings of the same species in nearby forest showed that leaf weight ratios and relative growth rates were all significantly lower in seedlings in the forest understorey (Bloor and Grubb 2003).Differences between forest seedlings and pot-grown seedlings may be partly due to the different nutrient availabilities in pots of different sizes.Soil volume restriction has effects on various plant responses that relate fundamentally to the reduction of photosynthesis per unit leaf area, based on a review of 65 studies by Poorter et al. (2012).In addition to differences due to soil nutrient availability, there may be differences due to light climate.Seedlings in the forest understorey experience variable light including sunflecks, whereas seedlings in greenhouses do not usually experience similar variations.
Of studies to date on the effect on tree seedlings of changing the light climate from understorey to gap light conditions, only two used seedlings in natural forest.The first, of eight species in Japanese temperate deciduous forest, showed that A max of pre-existing leaves of six species was higher 30 days post-gap compared to pre-gap and leaf mass per area was higher in three of these (Oguchi et al. 2006).In the second study, in the tropical lowland rainforest in Puerto Rico, only one of three species showed an increase in A max in preexisting leaves after a hurricane, the palm Prestoea acuminata (Wen et al. 2008).A third study grew plants in pots in the understorey of tropical forest in Panama and after two years moved the pots to a gap (Kursar and Coley 1999).One species (an Ouratea, with long-lived leaves) had increased A max about 46 days after transfer to the gap, and the other species (a Hybanthus with shorter-lived leaves) did not show increased A max .In addition to these three forest-based experiments, there are eight studies of tree seedling acclimation completed in shade houses or growth chambers.In these studies, there were a total of 27 species, of which 15 showed increased A max following the transfer from low light to higher light (Bauer and Thoni 1988, Strauss-Debenedetti and Bazzaz 1991, Kamaluddin and Grace 1992, Newell et al. 1993, Turnbull et al. 1993, Yamashita et al. 2000, Cai et al. 2005, Oguchi et al. 2005).
There is also interest in physiological aspects of the acclimation ability of invasive species, linked to various field observations of their ability to outcompete the native trees and shrubs in both gaps and shade.Pattison et al. (1998) found greater acclimation ability of invasive compared with native species in Hawaii, which was further supported by the study of invasive Melastomataceae by Baruch et al. (2000).Yamashita et al. (2000) noted differences in acclimation performance between pot-grown seedlings of tree species native to the Ogasawara Islands compared with the non-native, invasive Bischofia javanica.It acclimated more effectively than the native midsuccessional species with which it most closely competed due to the greater increase in A max of shade leaves after transfer, greater production of new leaves and increase in relative growth rate.A summary and analysis of these and other studies, by Fridley et al. (2023), indicate that forest invasions may be enhanced by the efficient carbon gain of the invasive species that can allow persistence and reproduction in the understorey, plus the ability of existing understorey plants to adapt to gaps faster than the native species.
Thus, it seems likely that many species, including some invasives, will have seedlings that will increase rates of photosynthesis in pre-existing leaves when these leaves are suddenly exposed to high light due to treefalls or defoliation of canopy trees after major windstorms.However, the acclimation of these 'shade-to-gap' leaves on seedlings growing in tropical rainforest, as a result of actual increases in light, has only been demonstrated once for Prestoea in lowland rainforest in Puerto Rico defoliated by a hurricane and that was for a palm that has exceptionally long-lived leaves compared to dicotyledonous trees (Wen et al. 2008).In that study, the leaves of two dicotyledonous species, whose seedlings were naturally grown in the shade, did not show acclimation (Wen et al. 2008).Our research investigated whether leaves of seedlings of six dicotyledonous species growing naturally in the understorey in a tropical montane rainforest in Jamaica would acclimate after the tree canopy over these seedlings was experimentally removed.

Study site and species
The forest lies just south of the crest of the Grand Ridge of the Blue Mountains in Jamaica at an altitude of 1570 m (Figure 1).The area is subject to frequent defoliating hurricanes.The site was in a 'Mull Ridge' forest type (Grubb andTanner 1976, Tanner 1977).Sixtyfive tree species are found in this forest.Our study focused on five common tree species and one shrub (Palicourea alpina).Two are light-demanders: Alchornea latifolia and Clethra occidentalis.Two are gap-favoured: Pittosporum undulatum (invasive) and Palicourea.One is intermediate: Hedyosmum arborescens, and one is shade-tolerant: Guarea glabra (Healey 1990, McDonald et al. 2003).Clethra has a tendency to establish on tree ferns (Newton and Healey 1989).Pittosporum undulatum is native to Victoria, Australia.It has been invading the forest for many decades and appears to be exacerbated by periodic hurricanes (Bellingham et al. 1995(Bellingham et al. , 2005;;Chai et al. 2012).
The six species are subsequently referred to by their generic names.Sizes of the seedlings and numbers of individuals in each treatment are in Table 1.Seedlings were all naturally occurring; potting and transplantation were not used in the experiment.

Canopy defoliation
Four gap and four adjacent understorey 10 × 10 m plots were established (Figure 1).All plots were south-facing with a slope angle of 10°-15°and were separated by a minimum of 40 m over ground.There was no sign of recent canopy disturbance and existing treefall, and other gaps were avoided.Within the four randomly chosen future gap plots, all leafy branches were removed from all trees over 5 m height.Additional branch removal along the solar arc outside the plots minimised diurnal interception of direct solar radiation by the canopy.The extra removal helped to simulate hurricane effects more accurately than a vertical-sided gap.The small size of the removed branches meant that they could be lowered to the ground carefully without damage to seedlings, and great care was taken to avoid such damage.
Photosynthetic photon flux density (PPFD) was measured in the open by a permanently installed sensor (Skye Instruments Ltd., Llandrindod Wells, UK), at Bellevue Peak, approximately 1.6 km east of the site.At ground level, it was measured by PPFD sensors made from silicon-blue photodiodes to the design of Biggs et al. (1971).These were calibrated against a manufacturer-calibrated sensor (Delta-T devices, Cambridge, UK) and found to have an accurate linear calibration (Ball 2001).A net of 30-35 sensors was placed randomly both before and after gap creation in the four gap plots and continuously in understorey plots.They were set to log every five seconds with one-minute integration, over a period of three to five days, which was statistically optimal (Ball 2001, Ch 2).PPFD in the gaps was between 22 and 39%, while that in the understorey was c. 3%, of the integrated diurnal PPFD above the canopy (Figure 2).Gaps were created in the following months: G1: January 1998; G2: February 1998; G3: April 1998; G4: September 1998.
Rainfall during the experiment was also measured at Bellevue Peak (Hafkenscheid 2000).The recorded rainfall was within 1 standard deviation of the mean rainfall at nearby Cinchona Gardens (1500 m) for each month (20 th -century average), lower than average from April to July 1998, higher than average during November 1998 and marginally lower than average during January to March 1999 (Figure 3).The dry spells did not induce any visible plant stress conditions at the site (T.Ball, personal observation).

Seedling measurements
Measurements were made on all seedlings of each species in each plot, except in the case of Guarea for which, due to its abundance in the understorey, 15 individuals on each plot were randomly selected for measurement (Table 1).If a seedling died, its existing data were eliminated from analyses; thus, the final data analyses were conducted only on the seedlings that survived the duration of the experiment.

Physiology
A max and stomatal conductance (g s ) were measured simultaneously in saturating light, at about 700 μmol/m 2 /s.An infrared gas analyser (IRGA) was used in open mode with a Parkinson leaf chamber attached (LCA-2 with PLC-B, Analytical Development Company [ADC], Hoddesdon, UK).Equipment was supported by field retort clamps adjustable to each plant.Seedlings in the understorey were illuminated with a 50-W halogen lamp from which interception of infrared energy was neutralised by a small water film sandwiched between Lexan plastic filters.This shield admitted 90% of incident PPFD; PPFD values (saturating light, in gaps c. 1100-1500 μmol m −2 s −1 ) were measured below the shield by placing a small quantum sensor in the leaf position.In the gap, shading via mesh material filters reduced light progressively, with the infrared filter film preserved.Leaf temperatures (calculated by the LCA-2 from its own radiometer and internal calculations) were 20-25°C in the understorey and 22-28°C in gaps.All chamber and lighting equipment was carefully mounted and aligned by adjusting the clamp position so that the leaf was kept at a natural angle.CO 2 at the chamber inlet was stabilised by a 5-litre mixing volume and kept within ± 10% of ambient levels (app.360-390 ppmv on the forest floor).Because this system had no chamber humidity control, a limited number of short replicate runs (generally 3-5) were carried out, keeping time in the chamber to approximately 40 seconds for each run.The minimum response time for this system is c. 30 seconds (ADC, LCA-2 manual).Measurements were taken in the morning hours between 8 a.m. and 12 noon, prior to any afternoon stomatal closure.R d was measured after darkening the whole seedling for 20 minutes using a box.These measurements were made on separate days from the A max and g s .For consistency, measurements were made on the first fully expanded, non-self-shaded leaf from the top of each plant.While these were closely assessed for full expansion and maturity, the extent of maturity may have differed between plants.Following an A max or R d measurement, no further measurements were taken on the leaf for several subsequent days.
We use the following nomenclature: leaves in the shade before gap treatment are called 'shade leaves', the shade leaves that were exposed in the gaps are called 'shade-to-gap leaves', and new leaves produced in the gaps are called 'gap leaves'.A max and g s values were measured pre-gap on shade leaves and repeated on the shade-togap leaves at intervals that were as consistent as possible after gap creation on each gap plot (in practice 4-6 days, 14-16 days, 35-38 days and 63-66 days and 130-137 days after gap creation).R d was measured pre-gap and at 24-28 and 82-87 days.In understorey plots, A max and g s were measured twice (pre-gap and post-gap dates, termed 'pre-control and 'post-control') and R d once (pregap).The central point of each time interval following was taken for plotting data.A max , g s and R d were measured on gap leaves once they had fully expanded, with expansion assessed by reference to existing mature leaves.

Leaf nitrogen concentration and leaf thickness
Following the final leaf census, the shade-to-gap leaf that had been measured for A max , along with the uppermost sun leaf (if fully expanded), was excised.The corresponding leaves in the understorey seedling groups were also sampled and excised in the same way.Four to six 5 mm diameter discs were removed from the intervenous area of the leaf and dried (65°C for 12 hours).The discs were then weighed to determine leaf dry mass per unit area (LMA).Leaf nitrogen concentration [N] was determined by the Kjeldahl method, using a Burkard scientific autoanalyser fitted with a colorimetric detector.LTs were measured with a dial micrometre, which did not compress the leaf, at six-week intervals.Measurements were taken at a minimum of six positions on either side of the midrib, between the side veins of the leaf.

Seedling growth and leaf demography
Seedling height was measured from ground level to the first fully expanded leaf from the top of the plant, before and six months after gap creation.Understorey plot seedling growth measurements were made at the start and end of 15 months of growth, in June to July 1997 and August to September 1998.For leaf demography, leaves were labelled using small wire loops on numerically tagged seedlings and recounted at intervals after gap creation.

Data analysis
Plot mean values were calculated for all leaf physiological variables (A max , R d and g s ) for each plot (gap and understorey).Individual species' mean values were then analysed for differences by the Mann-Whitney U (rank sum) test.Parametric tests were rejected owing to uncertainty about the normality of data distribution for the physiology data.These data sets were compared at two points in time: (1) at 4-6 days post-gap to pre-gap understorey and (2) gap measurement at 130-137 days post-gap (gaps) to five months in the understorey.LMA and LT data were compared between shade, shade-to-gap and gap leaves, again by Mann-Whitney U tests and also by regression of leaf [N] against shade, shade-to-gap and gap leaf A max .All differences reported in the results were significant at the 0.05 level unless otherwise stated.Leaf demography data were analysed for both the survivorship of the initial cohort present at gap creation and production of new leaves.Exponential fits were obtained to values of mean survivorship and mortality across all plots, and the exponential decay function was calculated according to the equations of Shiel et al. (1995).All statistical tests were performed using GENSTAT.

Maximum photosynthetic rates (A max ) and dark respiration (R d )
A reduction of A max was observed in shade-to-gap leaves in four of the six species (the other two showed non-significant reductions) at four to six days after gap creation (Figure 4, Table 2); the reduction was particularly marked for Pittosporum, at -65%.However, by approximately 4.5 months post-gap, the shade-to-gap leaves showed significantly elevated A max compared with their pregap levels in three species Alchornea (þ40%), Clethra (þ35%) and Palicourea (þ72%).Stomatal conductance was only significantly elevated in one of these, Palicourea (þ38%).Guarea, Pittosporum and Hedyosmum showed no significant photosynthetic acclimation of shade leaves.

Leaf thickness (LT) and leaf mass per unit area (LMA)
In gaps, significant increases were seen in the LT of shade-to-gap leaves in the three species that showed photosynthetic acclimation (Alchornea þ18%, Clethra þ18% and Palicourea þ14%; Figure 5).Only Alchornea and Pittosporum showed significant increases in LMA in shade-to-gap leaves (þ18% and þ 15%, respectively; Figure 6).

Leaf nitrogen concentrations
Leaf N concentrations ([N]) in understorey shade leaves ranged from 0.06 (Clethra) to 0.11 mg N cm −2 (Guarea, the most shadetolerant species; Figure 7).Shade-to-gap leaves showed significantly increased leaf [N] over pre-gap values in only one species (Clethra, elevated by 23%).Gap leaf nitrogen was significantly higher than shade leaf N in all six species (Figure 7).

Interrelationships between leaf characteristics and A max
A max increased significantly with leaf N per area across the range of leaf types and species (Figure 8a).However, there was no significant relationship within the set of shade and shade-to-gap data; the significance of the relationships noted was down to the high [N] of gap leaves.Similarly, across all leaf types, A max increased significantly with LT (Figure 8b), but the significance, as with the A max -N relationship, was due to the higher values recorded for gap leaves and was not significant within the combined shade and shade-to-gap data set.

Leaf demography
All species showed pronounced elevation in rates of leaf production and leaf loss in gaps compared with the understoreys (Table 3).Leaf production was considerably lower than mortality immediately after exposure to high light but later increased to levels above mortality (Figure 9).The total number of leaves was higher at four to five months post-gap than in the same plants before gap creation; in two of the light-demanding species that showed acclimation of A max (Alchornea 48% higher and Clethra 29% higher), increases in leaf numbers were lower for the other species (Guarea 11%, Hedyosmum 31%, Palicourea 8% and Pittosporum 14%).Pittosporum showed the lowest leaf mortality.Those seedlings with the lowest leaf production post-gap were also those that failed to increase their A max in shade-to-gap leaves.By the end of the experiment, gap leaves constituted 74% (Alchornea) and 71% (Clethra and Palicourea) of the total leaves but only 62% for Hedyosmum and 50% for Guarea and Pittosporum.

Seedling growth and survivorship
In broad terms, seedling growth paralleled indications of photosynthetic acclimation (shade-to-gap leaf acclimation, gap leaf production, gap leaf photosynthetic rate and leaf nutrient status).Growth was faster in species that showed greater acclimation (Table 1).The greatest relative height growth rates (RGRs) were seen in the introduced Pittosporum (24% in gaps and 7% in understorey) and the two most light-demanding native trees (Alchornea, RGR 16% gap, 3% understorey and Clethra, RGR 23% gap and 7% understorey).Intermediate were Palicourea (RGR 16% gap and 6% understorey) and Hedyosmum (RGR 9% gap and 5% understorey), while the shade-tolerant Guarea showed the slowest overall RGR (9% gap, 4% understorey; figures for all RGR derived from Table 1).Seedling survivorship in gaps was lower than in the understorey for all species, varying between 59 and 71% of all seedlings (Table 1).

Discussion
In the Jamaican montane rainforest, shade leaves on seedlings suddenly exposed to higher light due to canopy defoliation above them increased their A max in the two more light-demanding  T Ball and EV Tanner *Significance at 0.05 level.NS indicates not significant.Tests were conducted on plot average values.
Table 3. Leaf mortality coefficients (calculated from the equations of Shiel et al. 1995) and percentage survivorship and production (mean ± 1 standard error of the mean) of leaves at six months following gap creation.Also shown are the computed exponential decay coefficient (λ) and the median leaf lifespan, both calculated from the equations of Shiel et al. (1995).See Figure 9 for gap leaf production and mortality graphs  species (Alchornea and Clethra) and the gap-favouring Palicourea, and in one species, Clethra, there were also increases in leaf mass per area and leaf N. Two more shade-tolerant species (Guarea and Hedyosmum) and the gap-favoured Pittosporum had lower or unchanged A max in shade-to-gap leaves.This pattern of acclimation in shade leaves suddenly exposed to higher light was also seen in about half the species studied in other experiments.
In the Japanese warm temperate rainforest, six of eight studied species of naturally occurring tree seedlings increased their A max after sudden exposure to higher light following the creation of experimental gaps.The extent of the increase was similar among the six species and therefore unrelated to their relative light demand (Oguchi et al. 2017).In the semi-evergreen forest in Panama, of two shade-tolerant species, the one with long-lived leaves showed increased A max after exposure to a gap (Kursar and Coley 1999).In contrast, in a comparison of Australian rainforest species, the most light-demanding Omalanthus showed no acclimation when light levels were increased from 1% to 60%, whereas the mid-successional Duboisia and late successional Acmena did increase their A max (Turnbull et al. 1993).Thus, more light-demanding species often, but not always, show increased A max in shade leaves suddenly exposed to higher light.The increase in A max in shade-to-gap leaves in three Jamaican species was accompanied by increases in LT and in one species, Clethra, an increase in leaf mass per area and leaf N per area.In the only other study where leaf mass and leaf N were measured along with changes in A max in shade-to-gap leaves (Oguchi et al. 2006), four of the six species that showed an increase in A max also showed    T Ball and EV Tanner an increase in N, and three showed increases in leaf mass.Thus, it is common, but not universal, for increases in A max in shade-to-gap leaves to be accompanied by increases in LMA and sometimes leaf N.
Leaf production following exposure to gap light in the Jamaican seedlings was higher over six months in all six species, including the three species where there was no increase in A max in shade-tosun leaves (Table 3).In Japan, seedlings of four of five subtropical evergreen trees showed significantly higher new leaf production in shade-to-gap plants compared to those that remained in shade (over one month in plants growing in a shade house, Yamashita et al. 2000).In a different Japanese experiment with eight species (different species to those in the Yamashita et al. 2000) in experimental gaps in warm temperate rainforest, few leaves expanded or died in 31 days after gap creation (Oguchi et al. 2006).In a study of two species in Panama, seedlings of Hybanthus transferred from shade to a gap, lost most of their leaves and 80 days later had 4.9 times as many leaves compared to when transferred.The second species, Ouratea, retained most of its preexisting leaves and produced a few new leaves (Kursar and Coley 1999).Thus, seedlings of many species showed increased leaf production in shade plants suddenly exposed to gap light conditions; though the strongest and most rapid effects were seen in pot-grown plants in shade houses, naturally occurring seedlings in the forest showed no effect in a month after exposure to gap light (Oguchi et al. 2006) or took many months to produce new leaves (this study).
New leaf production was associated with increased A max in shade-to-gap leaves in our Jamaican study and in two other studies: in the five species of Japanese subtropical evergreen trees suddenly exposed to gap light conditions in a shade house experiment (Yamashita et al. 2000) and in the six species of subtropical trees and shrubs in a shade house experiment in China (Cai et al. 2005).In contrast, the two species of Panamanian understorey shrubs studied in natural forest shade and gaps show different responses: one made new leaves, and the other retained its pre-existing leaves and increased the A max in those leaves (Kursar and Coley 1999).Thus, many, but again, not all, species show both increased A max in preexisting leaves and increased production of new leaves in new gap light conditions.While the invasive Pittosporum did not show an increase in A max of shade leaves in this study, overall, it showed a high value for A max in both shade and sun leaves despite its very low leaf LMA and leaf nitrogen.Consequent photosynthetic nitrogen use efficiencies are very high for this species; this underlies the very strong performance of this species in the understorey as well as the high growth rates noted in the gaps, a feature of invasive species that reflects high efficiency in carbon economics (Fridley et al. 2023).
Comparisons of experiments with naturally growing seedlings compared to experiments with seedlings in shade houses are not strictly possible because no study has compared the same species in these different conditions.However, a review of the published experiments suggests that plants in pots in shade houses respond more quickly than naturally occurring shade-grown seedlings suddenly exposed to gap light conditions.Shade house experiments report changes in only a month, whereas experimental gap studies either report increases in A max or no increases in leaf production in a month (Oguchi et al. 2006) or a mixture of effects.A max increased in a few weeks, but leaf production effects took many months to manifest themselves.In this study in Jamaica, there were few noticeable differences between species in new leaf production by five weeks after exposure to gap light, but they became clear by 26 weeks.
Full acclimation to gap light conditions will only have occurred once all leaves on a seedling were produced in the new light conditions, because even in species that do show increased A max in pre-existing shade leaves, the increased A max is not as high as the A max in leaves produced in the gap light conditions (Figure 4 and Kursar and Coley 1999).In addition, pre-existing leaves survive for many monthsin Jamaica, 26 weeks after the creation of gaps, the six species retained on average 41 % (range 27-57%, Figure 9) of their original leaves.The loss of old leaves and production of new ones is a continuous rather than sudden process in seedlings of most species even if adult trees show a markedly seasonal pattern of leaf production and abscission (Aylett 1985).No study has monitored seedlings until all leaves were produced in the new gap light conditions.Our Jamaican study seems to be the longest, but even here, only about half the pre-existing leaves were lost after six months.It would probably require at least one year for all preexisting leaves to be lost (canopy leaves of large trees have an average lifespan of about one year [Aylett 1985]).In Panama, experiments would have needed to last many years since one of their species has leaves that typically last for five years (Kursar and Coley 1999).

Conclusion
Many, but by no means all, seedlings acclimate to being suddenly exposed to gap light conditions by increasing their A max and sometimes increasing their leaf mass and leaf N in pre-existing leaves.Seedlings also produce new leaves in the new gap light conditions.Such changes start soon after the increased light is imposed and, in naturally occurring seedlings, can take many months and will not be fully achieved until all leaves on a seedling are produced in the gap light conditions.Some species show no increase in A max in pre-existing leaves in the same conditions, whereas other species do show marked increases; one such species is the invasive alien Pittosporum in our Jamaican study, which only responds to a sudden increase in light by producing new leaves.Thus, seedlings growing in the low light of evergreen forest understoreys that are naturally exposed to much higher light after treefalls or wind defoliation of the canopy can respond by increasing photosynthesis in pre-existing leaves and by producing new leaves in the new light climate; acclimation of pre-existing leaves is presumably useful but is not necessary for a species to be successful.It is possible that a 'hump-shaped' response will eventually emerge when more species are investigated, with no increase in A max for very strongly light-demanding and strongly shade-tolerant species, but those of intermediate light demand will show increased A max in shade-to-gap leaves.In our study, the species that showed the most acclimation of A max in pre-existing leaves and lower rates of shade leaf survival were the more lightdemanding Alchornea, Clethra (canopy trees) and Palicourea (a subcanopy shrub); the three more shade-tolerant species, the introduced Pittosporum (which is successfully invading the native forests) and the native Hedyosmum and Guarea, showed no acclimation of shade leaves.

Figure 1 .
Figure 1.Study site and plot location.White band shows a 1500-m altitude contour.

Figure 2 .
Figure 2. Average levels of diurnal integrated photosynthetic photon flux density (PPFD) in the study plots following gap creation.Columns represent means (± one standard error) of a network of sensors (30-40), randomly distributed in each plot and averaged over days, related to an open-site sensor on a nearby hilltop.Gap data were collected two to six months after the creation of the gaps.PPFD values in the open are typically 25-30 mol/m 2 /d (Hafkenscheid 2000).

Figure 3 .
Figure 3. Rainfall near the study site (Bellevue Peak) during 1998 (open circles) and 1999 (filled circles).The 20 th -century averages are also shown (columns, error bars, 1 standard error) for Cinchona Botanical Gardens (data obtained from Jamaican Meteorological Service).

Figure 4 .
Figure 4. Leaf gas exchange characteristics during acclimation.Data points in the central part of each chart represent shade-to-gap leaves (present on seedlings before exposure of the leaves to high light).Data for leaves present on seedlings in understorey plots throughout the experiment are shown from pre-and post-gap samples.The right-hand part shows the data for fully expanded leaves that emerged after exposure to high light (gap leaves) with larger symbols.Point and error bars represent the mean and standard error across the study plots. 6

Figure 5 .
Figure 5. Leaf thickness of shade (black bars), shade-to-gap (grey bars) and gap (white bars) leaves at 150 days following gap creation.Shade leaves were taken from understorey plots.Asterisks indicate significant divergence from the species' shade leaf value (Mann-Whitney P > 0.05).

Figure 6 .
Figure 6.Leaf mass per unit area (LMA) at 150 days postgap.Notation for leaf type as per Figure 5.

Figure 7 .
Figure 7. Area-based N concentration of leaves sampled and preserved at 150 days post-gap.Notation for leaf type and significance as per Figures 5 and 6.
Figure 7. Area-based N concentration of leaves sampled and preserved at 150 days post-gap.Notation for leaf type and significance as per Figures 5 and 6. 8

Figure 8 .
Figure 8. a. (Left panel) relationship between measurements of area-based leaf nitrogen content [N] and maximum photosynthetic rates (A max ) for all leaf types at approximately five months following exposure to the gaps.Overall regression across all leaf types for all species is significant: A max = 16.08 [N] þ 0.32; F = 8.621, P = 0.097; R2 = 0.35; n = 18.b (Right panel) relationship between measurements of leaf thickness (LT) and maximum photosynthetic rates (A max ) for all leaf types at approximately five months following exposure to the gaps.Overall regression across all leaf types for all species is significant: A max = 0.013 (LT) þ 0.702; F = 9.39, P = 0.007; R2 = 0.37; n = 18.

Table 1 .
Size, growth, survivorship (over the course of the experiment) and sample size of seedlings sampled in the study plots *Seedlings defined as < 30cm in height, measured from soil to top node.Data ± 1 standard error of the mean in brackets.

Table 2 .
Two sample comparisons (using Mann-Whitney U tests) for photosynthetic and respiratory parameters collected at intervals following the creation of canopy gaps.Sign indicates elevation (þ) or depression (−) of the parameter compared to the control values